Spatiotemporal model improves survey indices for witch flounder stock assessment in the Grand Banks

Abstract

Accurate and precise estimates of survey abundance indices are essential inputs for stock assessment models and are important for the successful conservation and management of fisheries stocks. Since abundance indices can be standardized in various different ways, from conventional design-based approaches to model-based approaches, it is essential to compare the efficiency of those approaches and quantify the consequences for stock assessments. In this study, we focus on an important commercial stock of witch flounder (Glyptocephalus cynoglossus) in NAFO 3N+3O divisions. We first compare the survey indices standardized by design-based and model-based approaches, and find that a model-based estimator provides more precise estimates of survey indices than design-based approach for this stock. We then apply a widely used age-structured statistical catch-at-length (ACL) assessment model to the standardized catch-at-length survey indices from both approaches to estimate the age-based population dynamics for this stock. We conclude that the ACL model fit to model-based indices performed better than the same model fit to design-based indices.